I built on Elementor for six years. I sold Elementor builds to dozens of Bergen County clients. I have a Pro license that auto-renews. And in 2025, I stopped recommending it for any site where performance actually matters. Here’s the honest reasoning, including when I still think it’s the right tool.
The catalyst was a side-by-side test. Same client, same content, same hosting, same images. Elementor build: 287KB of CSS, 412KB of JavaScript, 3.8s LCP on 4G mobile, Lighthouse mobile score 47. Gutenberg + GeneratePress rebuild of the exact same pages: 41KB of CSS, 89KB of JavaScript, 1.2s LCP, Lighthouse 96. Same content. Same designer. Same coffee budget. The Elementor tax was real and it was costing the client roughly 12% of their organic traffic according to Search Console data we tracked for the 90 days after the swap.
The Bloat Tax Is Structural, Not Tunable
Elementor’s architecture is a render-time interpreter. Every section, column, and widget on your page is described in serialized JSON in _elementor_data, and Elementor parses that JSON on every request, generates HTML, generates CSS for that specific page, and ships the whole thing to the browser. Add jQuery dependency. Add Swiper. Add Lottie. Add motion effects engine. Add Elementor’s own animation library. Even on a 4-paragraph “About” page, the page weight starts at ~280KB before your content.
You can disable the bundled Google Fonts. You can turn off motion effects. You can flip Improved CSS Loading and Improved Asset Loading. I have done all of those. The floor on a tuned Elementor site is still ~3x heavier than a tuned Gutenberg site. The bloat is structural — it’s the cost of the interpreter pattern. No setting tweak removes it.
When Elementor Is Still the Right Tool
I’m not saying nobody should use it. I’m saying use it deliberately. These are the cases where it still wins:
- Non-technical owners who need to edit pages weekly and won’t touch Gutenberg block patterns.
- Visual-heavy marketing sites where conversion lift from polished design beats the SEO drag from bloat (think wedding venues, photographers, high-end real estate).
- Internal portals or intranets where SEO doesn’t matter and page weight is irrelevant.
- Quick-turn pages for events or campaigns with a 60-day lifespan — not worth a custom build.
- Clients already trained on Elementor where the retraining cost exceeds the performance gain.
If your B2B site lives or dies on organic traffic, none of those apply to you.
What I Use Instead in 2026
Three stacks, picked by the client’s technical comfort and budget:
- Gutenberg + GeneratePress Premium ($59/year): 90% of B2B builds. Native block editor, no proprietary builder lock-in, blazing fast, full theme.json design system control. Lighthouse 95+ out of the box.
- Gutenberg + Kadence Pro ($129/year): When the client wants more visual handholding in the editor. Heavier than GeneratePress but still 3-5x lighter than Elementor. Block library is excellent for non-developers.
- Custom block theme (FSE): For clients who can support a small annual maintenance budget and want zero theme dependency. Full Site Editing with curated patterns. The leanest possible footprint — sub-30KB CSS budget.
All three give the client a real block editor experience (not a proprietary builder that locks their content into a vendor format), real Lighthouse scores, and a clean upgrade path. Switching themes inside this stack is a one-day project. Switching off Elementor is a two-to-four week rebuild because every page is encoded in Elementor’s serialized format.
Real Performance Comparison From Last Quarter
I pulled Lighthouse data from 11 client sites we either built or rebuilt in Q1 2026. Median mobile scores by stack:
- Elementor + Hello theme + standard optimization: 52
- Elementor + Hello + FlyingPress + image CDN: 71
- Gutenberg + Kadence Pro: 89
- Gutenberg + GeneratePress Premium: 94
- Custom FSE block theme: 98
Even a heavily optimized Elementor site, with $200/year in extra caching plugins, didn’t catch a default Gutenberg + GeneratePress build with zero optimization plugins.
How AJD handles this
Every new build starts with a stack conversation, not a “we use Elementor” assumption. We ask about your traffic profile, who edits content, and what your performance baseline needs to be. Most B2B clients in Bergen County end up on Gutenberg + GeneratePress Premium with custom block patterns we build for their site. For existing Elementor clients with performance problems, we do an ROI analysis on a rebuild vs. an optimization pass — sometimes the optimization gets you to 75 and that’s enough. Sometimes the rebuild is the only way to 95. We tell you straight which one your situation calls for. Whether you work with us or not — if performance matters to your business, audit your current stack honestly.
Stuck on Elementor and watching your rankings slip? We’ll audit your current setup, show you the actual performance gap, and tell you whether to optimize or rebuild — no upsell pressure. Book Free Discovery Call →
